# Infrared spectroscopy to find water absorption lines

Tags:
1. Mar 29, 2015

### bertopolis

Hi I am currently trying to find the water absorption lines by using a Michelson interferometer, as a detector I am using an ocean spectrometer. The data obtained is thus the spectrum's "received" by the spectrometer. Am I right to assume that in order to find the absorption peak/peaks I should integrate the spectrum's obtained (between 700 and 1000nm), the plot of these would be a signal graph, which once this is Fourier transformed should give me the absorption lines? I am currently doing this using python and the results obtained are very misleading and unclear. The Fourier Transform's obtained will be attached to this post.
Thanks for any help

#### Attached Files:

File size:
4.2 KB
Views:
81
File size:
6.1 KB
Views:
79
• ###### otherfirst.png
File size:
3.9 KB
Views:
108
2. Mar 30, 2015

### blue_leaf77

I'm not familiar with the method you described there, did you read this from certain literature or were you told by someone?
Anyway why not using Beer-Lambert's law? From this you can get absorbance.

3. Mar 30, 2015

### bertopolis

I did this due to equipment limitations, I haven't heard of this law, and I don't exactly understand how it works. Could you please explain this?

4. Mar 30, 2015

### blue_leaf77

You have basically everything to make use of this equation because what you need is just a spectrometer. Beer-Lambert law states that $I(\omega,L) = I(\omega,0) e^{-a(\lambda)L}$ where $I(\omega,L)$ and $I(\omega,0)$ are the power spectrum (spectrometer signal) of the output and input light beam respectively, and L is the medium length (in this case the distance traveled in the water). The absorbance is defined as $a(\lambda)L$. So if you know both power spectra you can get the absorbance.

However remember that what you get from that calculation is the absorbance, which slightly different from absorption spectrum. They are equal only for small absorption, but if what you interested in is just the absorption peaks, this method should work.

Last edited: Mar 30, 2015
5. Mar 30, 2015

### bertopolis

That's brilliant and extremely helpful, thanks. SO does that mean the values I have from the spectrometer is I(w,z)-I(w,0) ? Is there anyway I could reach this equation from the values I have already obtained?

6. Mar 30, 2015

### blue_leaf77

You get I(w,L) from measuring the spectrum of the light you are using after it has passed through the water. While I(w,0) is the spectrum of your light source itself, that is the spectrum you get by placing the spectrometer directly in front your light source, with nothing in between. From this $I(\omega,L) = I(\omega,0) e^{-a(\lambda)L}$, I suppose you know how to extract $a(\lambda)L$.
I don't know how strong the absorption of water within infrared region is, but in case you notice that the absorption is too weak to observe you can try to increase L.